LiDAR Processing
Heightcraft provides robust support for processing LiDAR point clouds (.las, .laz). It is designed to handle large datasets efficiently using a streaming, chunked pipeline.
Features
- Format Support:
.lasand.laz(compressed LAS). - Streaming: Processes files chunk-by-chunk to keep memory usage low.
- Auto-Detection: Automatically detects LiDAR files based on extension.
- 32-bit Output: Supports 32-bit Float TIFF for maximum precision.
Usage
Simply provide a .las or .laz file as input.
heightcraft scan.laz --output_path dem.tiff --bit_depth 32 --max_resolution 2048
Memory Management
For extremely large files (multi-gigabyte), Heightcraft automatically uses a chunked approach. You can control the chunk size if needed, though the default is usually optimal.
# Process in chunks of 5 million points
heightcraft scan.laz --chunk_size 5000000
Integration with Upscaling
You can combine LiDAR processing with AI upscaling to get ultra-high-resolution DEMs from lower-density point clouds.
heightcraft scan.laz --upscale --upscale_factor 2 --output_path high_res_dem.tiff